strisunshinewentingwang/pytorch

Pull Requests Commits

Autograd graphtask trim unnecessary edges (#82544)

yueyericardo committed 2 years ago

382ef1fd

Add assertions to fix torch::jit::load bugs (#79192)

apach301 committed 2 years ago

d438e867

When encountering dynamic types, one should cast it recursively. (#83218)

qihqi committed 2 years ago

2ae1afd6

Fix issues with Werror=range-loop-construct (#83273)

seemethere committed 2 years ago

e88dea0c

[quant][fx] Add more detailed docs for prepare_fx/prepare_qat_fx/convert_fx (#83132)

jerryzh168 committed 2 years ago

bce1540f

Limits constant chunk propagation for pw-node-only (#83083)

jjsjann123 committed 2 years ago

a395f6e8

Fix TORCH_CHECK macros when glog is used (#83216)

wconstab committed 2 years ago

abb2204f

Change the input of `mvit_v2_s` on the FX test (#83242)

datumbox committed 2 years ago

cff55682

Hack up make_fx to natively support varargs (#83210)

ezyang committed 2 years ago

63f35f1a

[mergebot] Update merge bot messages to explain flags (#82907)

zengk95 committed 2 years ago

23b5fcaa

[functorch] Add some more view+inplace grad+vmap tests (#83176)

zou3519 committed 2 years ago

fa54021a

[functorch] vjpvjp inplace testing (#83119)

zou3519 committed 2 years ago

f8c408b7

[functorch] in-place testing for test_vjp (#83114)

zou3519 committed 2 years ago

ffc4a502

[functorch] in-place jvp testing (#83077)

zou3519 committed 2 years ago

3dc402fd

Propose code owners from Intel (#80218)

jgong5 committed 2 years ago

beceb8b9

Add nvfuser support for prims.sign and refs.sign (#83167)

IvanYashchuk committed 2 years ago

7191ae58

[xla hash update] update the pinned xla hash (#83236)

pytorchmergebot committed 2 years ago

c35db967

CUDA trace Python hooks (#82824)

sypneiwski committed 2 years ago

916def84

Fix broken FX tests (#83187)

datumbox committed 2 years ago

6a09847c

[PyTorch] Solved two syntax issues when dumping execution graph result to json file. (#81854)

mingyul-fb committed 2 years ago

ccfbbc0f

Preserve node's stack trace during retrace (#83050)

SherlockNoMad committed 2 years ago

69156764

[functorch] remove some testing hacks (#83079)

zou3519 committed 2 years ago

3aeb5e4f

Move vmap's OrderedDict pytree support to torch.utils._pytree (#83073)

zou3519 committed 2 years ago

6700a785

Prefer contiguous output from mkldnn_bf16_gemm (#82968)

peterbell10 committed 2 years ago

7d89c3b0

Handle redispatch correctly with tensor subclasses in ProxyTensor mode (#83122)

Chillee committed 2 years ago

66396777

add 3 operators ops (max_pool2d, linear, conv2d)in xnnPack test (#83131)

Cuiqing Li committed 2 years ago

22830c7c

[ROCm] update nightly builds to rocm5.2 (#82353)

jeffdaily committed 2 years ago

2dae93b2

[caffe2] Make ListIterator compatible with libc++ 15 (#83189)

smeenai committed 2 years ago

9590cf6d

[fx] Implement __deepcopy__ for fx.Tracer (#83130)

davidberard98 committed 2 years ago

45e7d026

[ao] Create framework for ModelReport Qconfig Generation (#83091)

vspenubarthi committed 2 years ago

a2ca8933

Newer Older